

** 10.12.2024
** Making time to be informed: The child penalty in political news consumption
** Dataset - LISS 


********************************************************************************

cd ""

/* Dataset used */ use "Pooled LISS data 1-15_replication file", clear

********************************************************************************

** Analyses 

/* Identify the dataset as panel-structured */ xtset nomem_encr wave

***# Figure 1. Gender differences in news consumption

/* Newspaper consumption */ 
ciplot po009 if wave<2020, by(gender) title("Newspapers (mean 95% CI)") note(" ") ytitle(" ") xtitle(" ") saving(fig01, replace)

 /* TV/radio consumption */
bys gender: egen mmediaTV=mean(mediaTV) if wave<2020
gen pmediaTV=mmediaTV*100

ciplot pmediaTV if wave<2020, by(gender) saving(fig02, replace) ytitle("") xtitle(" ") title("TV/Radio (percentage)") note(" ")

/* Online content consumption */
bys gender: egen mmediaWeb=mean(mediaWeb) if wave<2020
gen pmediaWeb=mmediaWeb*100

ciplot pmediaWeb if wave<2020, by(gender) saving(fig03, replace) ytitle("") xtitle(" ") title("Online (percentage)") note(" ")

graph combine fig01.gph fig02.gph fig03.gph, row(1)


**# Table 1. Fixed effects model. Threshold effect. 2008-2019.

* Newspapers
 /* Men */ xtreg po009 i.childrenalt if gender==0 & fa004<=50 & numchildren<2 & wave<2020, fe
 est store fix01a20h
 outreg2 using "tab1",  excel dec(2) pdec(3) replace
 /* Women */ xtreg po009 i.childrenalt if gender==1 & fa004<=50 & numchildren<2 & wave<2020, fe
 est store fix01a20m
  outreg2 using "tab1",  excel dec(2) pdec(3) append
  
* TV / radio 
  /* Men */ xtlogit mediaTV i.childrenalt if gender==0 & fa004<=50 & numchildren<2 & wave<2020, fe
 est store fix01b20h
   outreg2 using "tab1",  excel dec(2) pdec(3) append 
 /* Women */ xtlogit mediaTV i.childrenalt if gender==1 & fa004<=50 & numchildren<2 & wave<2020, fe
 est store fix01b20m
   outreg2 using "tab1",  excel dec(2) pdec(3) append

* Internet
   /* Men */ xtlogit mediaWeb i.childrenalt if gender==0 & fa004<=50 & numchildren<2 & wave<2020, fe
 est store fix01c20h
   outreg2 using "tab1",  excel dec(2) pdec(3) append
 /* Women */ xtlogit mediaWeb i.childrenalt if gender==1 & fa004<=50 & numchildren<2 & wave<2020, fe
 est store fix01c20m
   outreg2 using "tab1",  excel dec(2) pdec(3) append

   
**# Figure 2. Predicted media consumption for men and women when they have their first child (threshold hypothesis).

est restore fix01a20h
margins childrenalt, saving(fix01a20h, replace)
 est restore fix01a20m
margins childrenalt, saving(fix01a20m, replace)
combomarginsplot fix01a20h fix01a20m, recast(scatter) labels("Men" "Women") legend(cols(2) pos(6)) title("") saving(fig02, replace)


**# Table 2. Fixed effects models. Building up and short-term effects. 2008-2019.

* Newspaper
/* Men */ xtreg po009 ib1.parentcat if gender==0 & fa004<=50 & wave<2020, fe
 est store fix02a20h
   outreg2 using "tab2",  excel dec(2) pdec(3) append 
 /* Women */ xtreg po009 ib1.parentcat if gender==1 & fa004<=50 & wave<2020, fe
 est store fix02a20m
   outreg2 using "tab2",  excel dec(2) pdec(3) append
 
* TV/radio
  /* Men */ xtlogit mediaTV i.parentcat if gender==0 & fa004<=50 & wave<2020, fe
 est store fix02b20h
   outreg2 using "tab2",  excel dec(2) pdec(3) append
 /* Women */ xtlogit mediaTV i.parentcat if gender==1 & fa004<=50 & wave<2020, fe
 est store fix02b20m
   outreg2 using "tab2",  excel dec(2) pdec(3) append

* Internet   
   /* Men */ xtlogit mediaWeb i.parentcat if gender==0 & fa004<=50 & wave<2020, fe
 est store fix02c20h
   outreg2 using "tab2",  excel dec(2) pdec(3) append
 /* Women */ xtlogit mediaWeb i.parentcat if gender==1 & fa004<=50 & wave<2020, fe
 est store fix02c20m
   outreg2 using "tab2",  excel dec(2) pdec(3) append


**# Figure 3. Predicted media consumption for men and women over the age of their youngest child. Short term and building-up hypothesis.

est restore fix02a20h
 margins i.parentcat, saving(fix02a20h, replace)
 est restore fix02a20m
 margins i.parentcat, saving(fix0a20m, replace)
combomarginsplot fix02a20h fix0a20m, recast(scatter) labels("Men" "Women") legend(cols(2) pos(6)) title(" ") saving(fig03, replace)


**# Table 3. Fixed effects models. Extra burden effect. 2008-2019.

* Newspapers
 /* Men */ xtreg po009 i.childrenalt##i.active if fa004<=50 & gender==0 & wave<2020, fe
est store fix03a20h
  outreg2 using "tab3",  excel dec(2) pdec(3) append
/* Women */ xtreg po009 i.childrenalt##i.active if fa004<=50 & gender==1 & wave<2020, fe
est store fix03a20m
  outreg2 using "tab3",  excel dec(2) pdec(3) append

* TV/radio
 /* Men */ xtlogit mediaTV i.childrenalt##i.active if fa004<=50 & gender==0 & wave<2020, fe
est store fix03b20h
  outreg2 using "tab3",  excel dec(2) pdec(3) append
/* Women */ xtlogit mediaTV i.childrenalt##i.active if fa004<=50 & gender==1 & wave<2020, fe
est store fix03b20m
  outreg2 using "tab3",  excel dec(2) pdec(3) append

* Internet  
 /* Men */ xtlogit mediaWeb i.childrenalt##i.active if fa004<=50 & gender==0 & wave<2020, fe
est store fix03c20h
  outreg2 using "tab3",  excel dec(2) pdec(3) append
/* Women */ xtlogit mediaWeb i.childrenalt##i.active if fa004<=50 & gender==1 & wave<2020, fe
est store fix03c20m
  outreg2 using "tab3",  excel dec(2) pdec(3) append


**# Table 4. Fixed effects models. Fertility intentions. 2008-2019.

* Newspapers
/* Men*/ xtreg po009 ib2.fa128 if fa004<=50 & gender==0 & wave<2020,fe
est store fix04a20h
outreg2 using "tab4", excel dec(2) pdec(3) replace
/* Women */ xtreg po009 ib2.fa128 if fa004<=50 & gender==1 & wave<2020,fe
est store fix04a20m
outreg2 using "tab4", excel dec(2) pdec(3) append

* TV/Radio
/* Men*/ xtlogit mediaTV ib2.fa128 if fa004<=50 & gender==0 & wave<2020,fe
est store fix04b20h
outreg2 using "tab4", excel dec(2) pdec(3) append
/* Women */ xtlogit mediaTV ib2.fa128 if fa004<=50 & gender==1 & wave<2020,fe
est store fix04b20m
outreg2 using "tab4", excel dec(2) pdec(3) append

* Internet
/* Men*/ xtlogit mediaWeb ib2.fa128 if fa004<=50 & gender==0 & wave<2020,fe
est store fix04c20h
outreg2 using "tab4", excel dec(2) pdec(3) append
/* Women */ xtlogit mediaWeb i.b2.fa128 if fa004<=50 & gender==1 & wave<2020,fe
est store fix04c20m
outreg2 using "tab4", excel dec(2) pdec(3) append


********************************************************************************

**# ONLINE APPENDIX

**# Robustness checks

**# See below for replication of Contextual data, Figures A1 to A7. 

**# Table A2-A5. Descriptives of the dependent and independent variables.
tabout po009 mediaTV mediaWeb active childrenalt parentcat wave using "taba2.xls", c(col) f(1 1) stats(chi2) dpcomma replace

**# Table A6. Media consumption over wave and gender.
asdoc tab3way po009 gender wave, colpct nofreq save(taba6-1.doc)
asdoc tab3way mediaTV gender wave, colpct nofreq save(taba6-2.doc)
asdoc tab3way mediaWeb gender wave, colpct nofreq save(taba6-3.doc)

**# Table A7. Replication of tables 1 and 2 using free media consumption as dependent variables

* At least one child
   /* Men */ xtlogit mediafreenews i.childrenalt if gender==0 & fa004<=50 & numchildren<2 & wave<2020, fe
 outreg2 using "taba7", excel dec(2) pdec(3) replace 
 /* Women */ xtlogit mediafreenews i.childrenalt if gender==1 & fa004<=50 & numchildren<2 & wave<2020, fe
 outreg2 using "taba7", excel dec(2) pdec(3)

* Parent categories 
/* Men */ xtlogit mediafreenews i.parentcat if gender==0 & fa004<=50 & wave<2020, fe
 outreg2 using "taba7", excel dec(2) pdec(3)
 /* Women */ xtlogit mediafreenews i.parentcat if gender==1 & fa004<=50 & wave<2020, fe
 outreg2 using "taba7", excel dec(2) pdec(3)
 

**# Table A8. Average news consumption and differences for and between men and women.

* Data between 2008-2019
 ttest po009 if wave<2020, by(gender)
 ttest mediaTV if wave<2020, by(gender)
 ttest mediaWeb if wave<2020, by(gender)
 ttest po011 if wave<2020, by(gender)
 ttest mediafreenews if wave<2020, by(gender)
 ttest mediapaidnews if wave<2020, by(gender)
 
* Full dataset - 2008-2022
 ttest po009, by(gender)
 ttest mediaTV, by(gender)
 ttest mediaWeb, by(gender)
 ttest po011, by(gender)
 ttest mediafreenews, by(gender)
 ttest mediapaidnews, by(gender) 

 
**# Table A9. Replication of table 1 distinguishing between part-time and full-time employment.

*Newspaper
/* Men */ xtreg po009 i.childrenalt##i.typejob if fa004<=50 & gender==0 & wave<2020, fe
est store newsa91
outreg2 using "taba9", excel replace
/* Women */ xtreg po009 i.childrenalt##i.typejob if fa004<=50 & gender==1 & wave<2020, fe
est store newsa92
outreg2 using "taba9", excel append
   
* TV/radio 
/* Men */ xtlogit mediaTV i.childrenalt##i.typejob if fa004<=50 & gender==0 & wave<2020, fe
outreg2 using "taba9", excel append
/* Women */  xtlogit mediaTV i.childrenalt##i.typejob if fa004<=50 & gender==1 & wave<2020, fe
outreg2 using "taba9", excel append
  
* Online
/* Men */ xtlogit mediaWeb i.childrenalt##i.typejob if fa004<=50 & gender==0 & wave<2020, fe
outreg2 using "taba9", excel append
/* Women */  xtlogit mediaWeb i.childrenalt##i.typejob if fa004<=50 & gender==1 & wave<2020, fe
outreg2 using "taba9", excel append

  
**# Figure A9

/* Men */ est restore newsa91
margins typejob#childrenalt
marginsplot, xtitle("") title("Men") saving(parttimem.gph, replace)
/* Women*/ est restore newsa92
margins typejob#childrenalt
marginsplot,  xtitle("") title("Women") saving(parttimew.gph, replace)
/* Combination */ grc1leg2 parttimem.gph parttimew.gph, ycommon saving(figa9, replace) 


  
**# Table A10. Replication of tables 1 and 2 controling for "living with partner"

*** Newspapers
** Hypothesis 1
 /* Men */ xtreg po009 i.childrenalt partner if gender==0 & fa004<=50 & numchildren<2 & wave<2020, fe
  outreg2 using "taba10", excel dec(2) pdec(3) replace
 /* Women */ xtreg po009 i.childrenalt partner if gender==1 & fa004<=50 & numchildren<2 & wave<2020, fe
 outreg2 using "taba10", dec(2) pdec(3) excel
 ** Hypotheses 2 & 3.
 /* Men */ xtreg po009 i.parentcat partner if gender==0 & fa004<=50 & wave<2020, fe
 outreg2 using "taba10",  dec(2) pdec(3) excel
 /* Women */ xtreg po009 i.parentcat partner if gender==1 & fa004<=50 & wave<2020, fe
 outreg2 using "taba10", dec(2) pdec(3) excel
 
 ***TV
  ** Hypothesis 1
 /* Men */ xtlogit mediaTV i.childrenalt partner if gender==0 & fa004<=50 & numchildren<2 & wave<2020, fe
  outreg2 using "taba10", dec(2) pdec(3) excel append
 /* Women */ xtlogit mediaTV i.childrenalt partner if gender==1 & fa004<=50 & numchildren<2 & wave<2020, fe
 outreg2 using "taba10", dec(2) pdec(3) excel append
 ** Hypotheses 2 & 3.
 /* Men */ xtlogit mediaTV i.parentcat partner if gender==0 & fa004<=50 & wave<2020, fe
 outreg2 using "taba10", dec(2) pdec(3) excel append
 /* Women */ xtlogit mediaTV i.parentcat partner if gender==1 & fa004<=50 & wave<2020, fe
 outreg2 using "taba10", dec(2) pdec(3) excel append
 
 *** Internet
  ** Hypothesis 1
 /* Men */ xtlogit mediaWeb i.childrenalt partner if gender==0 & fa004<=50 & numchildren<2 & wave<2020, fe
  outreg2 using "taba10", dec(2) pdec(3) excel append
 /* Women */ xtlogit mediaWeb i.childrenalt partner if gender==1 & fa004<=50 & numchildren<2 & wave<2020, fe
 outreg2 using "taba10",  dec(2) pdec(3) excel append
 ** Hypotheses 2 & 3.
 /* Men */ xtlogit mediaWeb i.parentcat partner if gender==0 & fa004<=50 & wave<2020, fe
 outreg2 using "taba10", dec(2) pdec(3) excel append
 /* Women */ xtlogit mediaWeb i.parentcat partner if gender==1 & fa004<=50 & wave<2020, fe
 outreg2 using "taba10", dec(2) pdec(3) excel append
 

**# Table A11. Replication of models in Table 1 with other children at home.

* Newspapers
/* Men */ xtreg po009 i.parentcatCORR if gender==0 & fa004<=50 & wave<2020, fe
 outreg2 using "taba11", excel dec(2) pdec(3) replace
/* Women */ xtreg po009 i.parentcatCORR if gender==1 & fa004<=50 & wave<2020, fe
 outreg2 using "taba11", dec(2) pdec(3) excel
 
* TV/radio
/* Men */ xtlogit mediaTV i.parentcatCORR if gender==0 & fa004<=50 & wave<2020, fe
 outreg2 using "taba11", dec(2) pdec(3) excel 
/* Women */ xtlogit mediaTV i.parentcatCORR if gender==1 & fa004<=50 & wave<2020, fe
 outreg2 using "taba11", dec(2) pdec(3) excel
 
* Online
/* Men */ xtlogit mediaWeb i.parentcatCORR if gender==0 & fa004<=50 & wave<2020, fe
 outreg2 using "taba11", dec(2) pdec(3) excel 
/* Women */ xtlogit mediaWeb i.parentcatCORR if gender==1 & fa004<=50 & wave<2020, fe
 outreg2 using "taba11", dec(2) pdec(3) excel


**# Table A12. Replication of models in Table 1 controlling by parental leave, newspapers.

*** Newspapers
 ** Hypothesis 1
 /* Men */ xtreg po009 i.childrenalt parentleave if gender==0 & fa004<=50 &  numchildren<2 & wave<2020 , fe
  outreg2 using "taba12", dec(2) pdec(3) excel replace
 /* Women */ xtreg po009 i.childrenalt parentleave if gender==1 & fa004<=50 &  numchildren<2 & wave<2020, fe
 outreg2 using "taba12", dec(2) pdec(3) excel
 ** Hypotheses 2 & 3.
 /* Men */ xtreg po009 i.parentcat parentleave if gender==0 & fa004<=50 & wave<2020, fe
 outreg2 using "taba12", dec(2) pdec(3) excel
 /* Women */ xtreg po009 i.parentcat parentleave if gender==1 & fa004<=50 & wave<2020, fe
 outreg2 using "taba12", dec(2) pdec(3) excel
 
*** TV/radio and Internet are not included in the results due to small N
 
 
**# Table A13. Replication of models in Table 1 controlling for parents reducing their paid working hours to take care of children.

*** Newspapers
  ** Hypothesis 1
 /* Men */ xtreg po009 i.childrenalt workcare if gender==0 & fa004<=50 &  numchildren<2 & wave<2020, fe
  outreg2 using "taba13", dec(2) pdec(3) excel replace
 /* Women */ xtreg po009 i.childrenalt workcare if gender==1 & fa004<=50 &  numchildren<2 & wave<2020, fe
 outreg2 using "taba13", dec(2) pdec(3) excel
 ** Hypotheses 2 & 3.
 /* Men */ xtreg po009 i.parentcat workcare if gender==0 & fa004<=50 & wave<2020, fe
 outreg2 using "taba13", dec(2) pdec(3) excel
 /* Women */ xtreg po009 i.parentcat workcare if gender==1 & fa004<=50 & wave<2020, fe
 outreg2 using "taba13", dec(2) pdec(3) excel
 
*** TV/radio and Internet are not included in the results due to small N
 
**# Table A14. Replication of Table 2 in the main text only for respondents without children.

* Newspaper
/* Men */ xtreg po009 i.selfselection if gender==0 & fa004<=50 & wave<2020, fe
outreg2 using "taba14", dec(2) pdec(3) excel
/* Women */ xtreg po009 i.selfselection if gender==1 & fa004<=50 & wave<2020, fe
outreg2 using "taba14", dec(2) pdec(3) excel

* Television
/* Men */ xtlogit mediaTV i.selfselection if gender==0 & fa004<=50 & wave<2020, fe
outreg2 using "taba14", dec(2) pdec(3) excel
/* Women */ xtlogit mediaTV i.selfselection if gender==1 & fa004<=50 & wave<2020, fe
outreg2 using "taba14", dec(2) pdec(3) excel

* Internet
/* Men */ xtlogit mediaWeb i.selfselection if gender==0 & fa004<=50 & wave<2020, fe
outreg2 using "taba14", dec(2) pdec(3) excel
/* Women */ xtlogit mediaWeb i.selfselection if gender==1 & fa004<=50 & wave<2020, fe
outreg2 using "taba14", dec(2) pdec(3) excel



**# Table A15. Replication of table 1 for those with and without childcare.

* Newspaper
  /* Men */ xtreg po009  i.childcarewithnonparents if gender==0 & fa004<=50  & wave<2020 & numchildren<2, fe
outreg2 using "taba15", dec(2) pdec(3) excel
 /* Women */ xtreg po009  i.childcarewithnonparents if gender==1 & fa004<=50  & wave<2020 & numchildren<2, fe
outreg2 using "taba15", dec(2) pdec(3) excel replace

* Television 
  /* Men */ xtlogit mediaTV  i.childcarewithnonparents if gender==0 & fa004<=50  & wave<2020 & numchildren<2, fe
outreg2 using "taba15", dec(2) pdec(3) excel
 /* Women */ xtlogit mediaTV  i.childcarewithnonparents if gender==1 & fa004<=50  & wave<2020 & numchildren<2, fe
outreg2 using "taba15", dec(2) pdec(3) excel

 * Internet
 /* Men */ xtlogit mediaWeb  i.childcarewithnonparents if gender==0 & fa004<=50  & wave<2020 & numchildren<2, fe
outreg2 using "taba15", dec(2) pdec(3) excel
 /* Women */ xtlogit mediaWeb  i.childcarewithnonparents if gender==1 & fa004<=50  & wave<2020 & numchildren<2, fe
outreg2 using "taba15", dec(2) pdec(3) excel


**# Table A16. Replication of table 2 for respondents with and without childcare.

* Newspaper
/* Men */ xtreg po009 ib1.kidcare if gender==0 & fa004<=50 & wave<2020, fe
outreg2 using "taba16", dec(2) pdec(3) excel
/* Women */ xtreg po009 ib1.kidcare if gender==1 & fa004<=50 & wave<2020, fe
outreg2 using "taba16", dec(2) pdec(3) excel

* Television
/* Men */xtlogit mediaTV ib1.kidcare if gender==0 & fa004<=50 & wave<2020, fe
outreg2 using "taba16", dec(2) pdec(3) excel
/* Women */ xtlogit mediaTV ib1.kidcare if gender==1 & fa004<=50 & wave<2020, fe
outreg2 using "taba16", dec(2) pdec(3) excel

* Internet
/* Men */xtlogit mediaWeb ib1.kidcare if gender==0 & fa004<=50 & wave<2020, fe
outreg2 using "taba16", dec(2) pdec(3) excel
/* Women */ xtlogit mediaWeb ib1.kidcare if gender==1 & fa004<=50 & wave<2020, fe
outreg2 using "taba16", dec(2) pdec(3) excel


**# Table A17. Political participation and political news consumption.

* Demonstration
/* Newspaper */ xtlogit demonstration fa004 i.education i.gender##i.po009 if fa004<=50 & wave<2020, re
est store demons
outreg2 using "taba17", dec(2) pdec(3) excel
/* TV/radio */ xtlogit demonstration fa004 i.education i.gender##i.mediaTV if fa004<=50 & wave<2020, re
outreg2 using "taba17", dec(2) pdec(3) excel
/* Internet */ xtlogit demonstration fa004 i.education i.gender##i.mediaWeb if fa004<=50 & wave<2020, re
outreg2 using "taba17", dec(2) pdec(3) excel

* Contact a politician
/* Newspaper */ xtlogit contact fa004 i.education i.gender##i.po009 if fa004<=50 & wave<2020, re
outreg2 using "taba17", dec(2) pdec(3) excel
/* TV/radio */ xtlogit contact fa004 i.education i.gender##i.mediaTV if fa004<=50 & wave<2020, re
outreg2 using "taba17", dec(2) pdec(3) excel
/* Internet */ xtlogit contact fa004 i.education i.gender##i.mediaWeb if fa004<=50 & wave<2020, re
outreg2 using "taba17", dec(2) pdec(3) excel

* Participate in a government-organized public hearing
/* Newspaper */ xtlogit hearing fa004 i.education i.gender##i.po009 if fa004<=50 & wave<2020, re
outreg2 using "taba17", dec(2) pdec(3) excel
/* TV/radio */ xtlogit hearing fa004 i.education i.gender##i.mediaTV if fa004<=50 & wave<2020, re
outreg2 using "taba17", dec(2) pdec(3) excel
/* Internet */ xtlogit hearing fa004 i.education i.gender##i.mediaWeb if fa004<=50 & wave<2020, re
outreg2 using "taba17", dec(2) pdec(3) excel


**# Figure A10. Predictive margins of political news consumption for men and women's probabilities to participate in a public demonstration


xtlogit demonstration fa004 i.education i.gender##i.po009 if fa004<=50 & wave<2020, re
margins i.gender#i.po009
marginsplot, xdim(po009)



*************************************************************
**# Additional descriptives - contextual data and information
*************************************************************

* Figure A1. 1.	Figure A1. News about politics and current affairs, watching, reading or listening, in minutes, ESS wave 10, 2020-2022.

// Using all countries that participated in wave 10. Datasets ESS10SCe03_1 and  ESS10e03_2

use "ESS10-ESS10SC-subset.dta", clear

encode cntry, gen(country)
gen polint=4-polintr
recode gndr (1=0) (2=1), gen(woman)
lab var woman "Woman"

** Models without controls

forval i=1/31 {
di "`i' : `: label country `i''"
reg nwspol woman if country==`i' 
est store gap`i'
}

coefplot (gap*), drop(_cons) xline(0) asequation swapnames ///
eqrename(^gap(.*)$ = \1.country, regex) ///
title("Women coefficients (no controls)") ysize(4) xsize(6)

graph save essgapnocontrols.gph, replace

est clear

** Models with controls: education (eduyrs), age (agea), political interest (polint)

forval i=1/31 {
di "`i' : `: label country `i''"
reg nwspol woman c.agea c.eduyrs  if country==`i' 
est store gap`i'
}

coefplot (gap*), drop(_cons agea eduyrs) xline(0) asequation swapnames ///
eqrename(^gap(.*)$ = \1.country, regex) ///
title("Women coefficients (with controls)")

graph save essgapcontrols.gph, replace

*** Figure A1:

graph combine essgapnocontrols.gph essgapcontrols.gph, 


**** Contexual information from Reuters' data:
***********************************************

use "reuters2023.dta", clear

**# Figure A2. Frequency of accessing the news over gender and country. Reuters Institute's Digital News Report. 

est clear
gen accessnews=10-q1b_new if q1b_new<=10
tab accessnews, m
recode gender_int (1=0) (2=1), gen(woman)
lab var woman "Woman"

* Models without controls 

forval i=1/46 {
di "`i' : `: label country `i''"
reg accessnews woman if country==`i'
est store gap`i'
}

coefplot (gap*), drop(_cons) xline(0) asequation swapnames ///
eqrename(^gap(.*)$ = \1.country, regex) ///
title("Woman coefficients (no controls)") ysize(4) xsize(6)
graph save reutersnocontrols.gph, replace

est clear

* Models with controls: age, income 

forval i=1/31 {
di "`i' : `: label country `i''"
reg accessnews woman c.age_int i.hh_income  if country==`i' 
est store gap`i'
}

coefplot (gap*), drop(_cons age_int 2.hh_income 1.hh_income 3.hh_income) xline(0) asequation swapnames ///
eqrename(^gap(.*)$ = \1.country, regex) ///
title("Woman coefficients (with controls)")

graph save reuterscontrols.gph, replace

**# Figure 2:
graph combine reutersnocontrols.gph reuterscontrols.gph, 


*# Figure A3. Patterns of news consumption by channel and gender. Reuters Institute's Digital News Report. 

clonevar favsource=q4 if q4<=11

preserve

* Collapse data
collapse (mean) q301-q313 if country<=46, by(gender)

lab var q301 "Television" 
lab var q302 "24 hour news television channels"
lab var q303 "Radio news"
lab var q304 "Printed Newspapers"
lab var q305 "Printed Magazines"
lab var q306 "Websites/apps of Newspapers"
lab var q307 "Websites/apps of news magazines"
lab var q308 "Websites/apps of TV and Radio companies"
lab var q309 "Websites/apps of other news outlets"
lab var q310 "Social media"
lab var q313 "None of these"

statplot q301-q313, over(gender_int)  asyvars bar(1, color(gray)) bar(2, color(blue)) recast(bar) ytitle("") saving(reuterspool.gph, replace) ytitle("")  

restore 

* Figure A4. Consumption patterns per country. Reuters Institute's Digital News Report. 

preserve


* Collapse data
collapse (mean) q301-q313, by(country gender)

lab var q301 "TV" 
lab var q302 "24-h news TV"
lab var q303 "Radio"
lab var q304 "Print Newspapers"
lab var q305 "Print Magazines"
lab var q306 "Web/app Newspapers"
lab var q307 "Web/app Magazines"
lab var q308 "Web/app TV/Radio"
lab var q309 "Web/app other"
lab var q310 "Social media"
lab var q313 "None of these"


decode country, gen(country_label)
sort country_label
gen new_order = _n
encode country_label, gen(new_country_var)
list new_country_var country_label, sep(0)

statplot q301-q313 if new_country_var<=8, over(gender_int)  asyvars bar(1, color(gray)) bar(2, color(blue)) recast(bar) ytitle("") by(new_country_var, rows(2) legend(off)) saving(country1.gph, replace) ytitle("") legend(off) 

statplot q301-q313 if new_country_var>=9 & new_country_var<=16, over(gender_int)  asyvars bar(1, color(gray)) bar(2, color(blue)) recast(bar) ytitle("") by(new_country_var, rows(2) legend(off)) saving(country2.gph, replace)

statplot q301-q313 if new_country_var>=17 & new_country_var<=24, over(gender_int)  asyvars bar(1, color(gray)) bar(2, color(blue)) recast(bar) ytitle("") by(new_country_var, rows(2) legend(off))  saving(country3.gph, replace)

statplot q301-q313 if new_country_var>=25 & new_country_var<=32, over(gender_int)  asyvars bar(1, color(gray)) bar(2, color(blue)) recast(bar) ytitle("") by(new_country_var, rows(2) legend(off)) saving(country4.gph, replace)

statplot q301-q313 if new_country_var>=33 & new_country_var<=40, over(gender_int)  asyvars bar(1, color(gray)) bar(2, color(blue)) recast(bar) ytitle("") by(new_country_var, rows(2) legend(off)) saving(country5.gph, replace)

statplot q301-q313 if new_country_var>=41 & new_country_var<=46, over(gender_int)  asyvars bar(1, color(gray)) bar(2, color(blue)) recast(bar) ytitle("") by(new_country_var, rows(2)) saving(country6.gph, replace)

*# Figure A5. Favourite source of news. Reuters Institute's Digital News Report. 

restore 

*clonevar favsource=q4 if q4<=11

lab define favsource 1 "TV" 2 "24h TV channels" 3 "Radio" 4 "Newspapers" 5"Magazines" 6 "Web/apps newspapers" 7 "Web/apps magazines" 8 "Web/apps TV & Radio" 9"Web/apps other" 10 "Social media" 11"None of these"
lab val favsource favsource
	
tabplot gender_int favsource,  percent(gender_int)	xtitle("") showval ytitle("")


*# Figure A6. Evolution of main news consumption channels over time. Reuters Institute's Digital News Report. 
use "reuters2013-2022.dta", clear

clonevar favsource=q4 if q4<=11

* Handeling missing data points. Data for years 2016 to 2022 only for countries up to 26th.

tab favsource year 
  
drop if country>=26
drop if year==2014
 
recode favsource (1/2=1) (3=2) (4/5=3) (6/7=4) (10=5), gen(favreduced)

lab define favreduced 1"TV" 2"Radio" 3"Printed newspapers/magazines" 4"Online newspapers/magazines" 5"Social Media"
lab val favreduced favreduced 


gen TVt=0
replace TVt=1 if q301==1 | q302==1
gen Radiot=0
replace Radiot=1 if q303==1 
gen Newspaperst=0
replace Newspaperst=1 if q304==1 | q306==1
gen SocialMediat=0
replace SocialMediat=1 if q310==1

egen mTV=mean(TVt), by(year)
egen mRadio=mean(Radiot), by(year)
egen mNewspapers=mean(Newspaperst), by(year)
egen mSocialMedia=mean(SocialMediat), by(year)

twoway connected mTV year || connected mRadio year || connected mNewspapers year || connected mSocialMedia year, legend(label(1 "TV") label( 2 "Radio") label(3 "Newspapers") label(4 "Social media"))

* Figure A7. Evolution of news consumption over time for men and women. Reuters Institute's Digital News Report. 

egen TV=mean(TVt), by(year gender_int)
egen Radio=mean(Radiot), by(year gender_int)
egen Newspapers=mean(Newspaperst), by(year gender_int)
egen SocialMedia=mean(SocialMediat), by(year gender_int)
 

twoway connected TV year if gender_int==1 || connected TV year if gender_int==2 || connected Radio year if gender_int==1 || connected Radio year if gender_int==2 || connected Newspapers year if gender_int==1 || connected Newspapers year if gender_int==2 || connected SocialMedia year if gender_int==1 || connected SocialMedia year if gender_int==2, ///
legend(label(1 "TV men") label(2 "TV women") label(3 "Radio men") label(4 "Radio women") label(5 "Newspapers men") label(6 "Newspapers women") label(7 "Social media men") label(8 "Social media women"))


* Figure A8. Evolution of news consumption over time for men and women in The Netherlands. Reuters Institute's Digital News Report. 

fre country
drop if country!=18


egen TVne=mean(TVt), by(year gender_int)
egen Radione=mean(Radiot), by(year gender_int)
egen Newspapersne=mean(Newspaperst), by(year gender_int)
egen SocialMediane=mean(SocialMediat), by(year gender_int)
 
twoway connected TVne year if gender_int==1 || connected TVne year if gender_int==2 || connected Radione year if gender_int==1 || connected Radione year if gender_int==2 || connected Newspapersne year if gender_int==1 || connected Newspapersne year if gender_int==2 || connected SocialMediane year if gender_int==1 || connected SocialMediane year if gender_int==2, ///
legend(label(1 "TV men") label(2 "TV women") label(3 "Radio men") label(4 "Radio women") label(5 "Newspapers men") label(6 "Newspapers women") label(7 "Social media men") label(8 "Social media women"))


